This study analyzes whether there is a threshold effect in the innovation-growth relationship. Using data from the period 2008–2017, we perform an analysis using 60 countries anel smooth transition regression (PSTR)in the whole sample and a split-sample analysis in which we separate developed countries (36) from developing countries (24). The results for the panel smooth transition regression (PSTR) model indicate that there is a threshold effect in the innovation-growth relationship. We find that below the threshold, the effect of innovation measured by the number of patents is not significant for developed and developing countries. However, surpassing the optimal threshold, the effect becomes positive only for the whole sample and developed countries. Furthermore, findings also indicate that research and development expenditure, domestic and foreign investments stimulate economic growth.